Histo-Clinical Classification and Follow-Up Study of Gastric Polyp

Transcription

1 Tohoku J. exp. Med., 1976, 118 (Suppl.), Histo-Clinical Classification and Follow-Up Study of Gastric Polyp KOTARO UENO,* SABUR OSHIBA, SHOICHI YAMAGATA, FUKUJI MOCHIZUKI, MASANOBU KITAGAWA and SHIGERU HISAMICHI The Third Department of Internal Medicine, Tohoku University School of Medicine, Sendai UENO, K., OSHIBA, S., YAMAGATA, S., MocmzuKI, F., KITAGAWA, M. and HIsAMICHI, S. Histo-Clinical Classification and Follow-Up Study of Gastric Polyp. Tohoku J. exp. Med., 1976, 118 (Suppl.), Gastric polyp was found in 1,616 cases or 0.23% of 711,455 persons through the gastric mass-survey. They were classified into 3 groups; hyperplastic polyp in 61%, gastritis polyposa in 34%, and polyp composed of metaplastic epithelium with marked atypism (ATP) in 5%. Gastritis polyposa was subclassified into 5 groups according to its histo clinical features. Malignant degeneration was suspected in 5 or 2.1% of 236 hyperplastic polyps operated, while no polyp which had changed into cancer was experienced among 974 polyps followed-up by biopsy during the time course of 6 months to 11 years at the longest. On the other hand, coexistence of gastric cancer was found in 13% of cases with ATP. In polyps composed of metaplastic epithelium, there are two kinds of polyp: metaplastic polyp in hyper plastic proliferation and neoplastic polyp. In the follow-up study of 1,104 polyps, the growth of polyp was seen in only 14, and 12 of which were hyperplastic polyp. No change from gastritis polyposa to hyperplastic polyp was observed. The increase in size in 14 polyps was seen mostly 3 to 4 years after they were found and was always transient. From these results, it is considered that most of gastric polyps clinically detected have already completed their growth. gastric polyp Although most of gastric polyps are thought to be benign, the malignant potential of polyps over 2 cm in size or of those composed of more primitive metaplastic epithelia continues to be a subject of considerable discussion (Castleman and Krickstein 1960). To this subject, it will be an effective aid to know the actual incidence of gastric polyp among people who never come under clinical observation because of no complaints, and to know how each type of these gastric polyps develops and how many of them may actually turn to malignancy in their course. The purpose of this paper is to describe the results of the clinico-histological observation of gastric polyp detected in the gastric mass-survey from the above point of view. MATERIALS AND METHODS During the past 11 years, a mass-survey for gastric diseases has been conducted on a total of 711,455 persons in a prefectural district population of about 2,000,000. The Received for publication, November 20, * Present address: The Second Department of Internal Medicine, Yamagata University School of Medicine, Yamagata. 23

2 24 K. Ueno et al. screening test by means of indirect fluorophotography was primarily done in all of the subjects, and gastrocamera or direct x-ray examination was then indicated for further study in about 20% of them. In cases in which some abnormality was found, fiber -scopic examination and biopsy were done to make definite diagnosis. Consequently, gastric polyp, which is defined as a benign discrete mucosal elevation due to epithelial proliferation, was found in 1,616 cases. Inflammatory fibroid polyp and submucosal tumors were excluded from the discussion. The gross appearance of gastric polyps was observed endoscopically, and the size of the lesion was measured indirectly by using FGS-BLG (Maruyama et al. 1972) or a scale attached to the biopsy forceps and by x-ray picture if necessary. In 1,104 polyps, the follow-up observation of the polyp in the clinical course has been done by fiberscopic examination and by biopsy during the time course of 6 months to 11 years at the longest. The interval between examinations was mostly 3 to 6 months at the beginning, and once or twice in a year in the later course. To study the relationship between gastric polyp and cancer, histologic specimens of 236 polyps and 61 polypoid cancers, which had been detected through the mass-survey and operated, were reviewed. Histologic specimens taken from polyps by biopsy or polypectomy were stained with hematoxylin-eosin and the multiple staining of alcian blue-pas-masson's trichrome method in order to study the mucin productivity of the epithelial cells composing polyp. Cell proliferation kinetics of the mucosa of some polyps was also investigated by means of in vitro autoradiography using 3H-TDR. RESULTS Histo-clinical types of polyps detected Gastric polyps so far diagnosed are classified as shown intable 1, according to the endoscopic findings and histologic findings of the biopsy or resected specimens. TABLE 1. Histologic types of gastric polyps Hyperplastic polyp is the commonest type seen in 61% of all polyps, and corresponds to that previously called adenomatous polyp by Stout (1953), being mainly composed of large columnar epithelial cells which contained much mucin in the cytoplasm and formed cystic dilated glands intermingled with increased collagen and smooth muscle fibers originating from muscularis mucosae. This type consisted of sessile polyps as well as pedunculated polyps, and the surrounding mucosa of these polyps was mostly atrophic. Hyperplastic polyp is at present found most definitely not to be tumorous in nature, so that they are also called regenerative polyp (Ming and Goldman 1965; Evans 1966). Among other polyps summarized as gastritis polyposa, five kinds are distinguished histologically. Type A is a polyp entirely consisting of normal fundic

3 Gastric Polyp 25 Fig. 1. Type A of gastritis polyposa. (a) Gastrocamera (GTF) picture of the polyp on the greater curvature of the upper portion of the corpus. (b) Its histologic picture consisting of normal fundic glands. glands, rarely seen around the greater curvature of the gastric corpus, being usually solitary, sessile and small in size as shown in Fig. 1. Polyps of types B and C are due to focal hyperplasia of the gastric epithelium caused by erosive gastritis. This type was termed gastritis verrucosa by Abel (1954). The gross appearance of this type shows a characteristic feature like suckers of octopus, having a small depression due to an erosion on the top of the mucosal elevation. Type B is a polyp predominantly consisting of pseudopyloric glands in hyperplastic prolifera tion, showing rosary-like verrucous appearance, with characteristic multiple occur rence in the antrum (Fig. 2). Type C, on the contrary, mainly consists of regenera tive epithelium of the gastric foveolae in hyperplastic proliferation, with characteristic multiple occurrence as polyposis on the portion of the fundic mucosa close to the border-line between the pyloric gland area and fundic gland area (Fig. 3). In 9.5% of the cases with gastritis verrucosa, pedunculated hyperplastic polyp was found concomitantly in the same stomach. The surrounding gastric mucosa of these polyps examined by biopsy was less atrophic. The gastric juice analysis revealed normo or hyperacidity of gastric juice in 65% of the cases with type B and 12% of type C. In type D, no characteristic findings of polyp but solely gastritis were observed at gastric biopsy. Most of these consisted of small sessile polyps. Type E represents polyps with elevation around the lymph follicles within the gastric mucosa, rarely seen in the antrum as multiple polyposis. Most of these are sessile and small in size, but some are long and pedunculated, in which

4 26 K. Ueno et al. Fig. 2. Type B of gastritis polyposa. (a) Gastrocamera (GTF) picture showing large and small verrucous protrusions in the antrum. (b) Histologic picture of the large ver rucous protrusion predominantly consisting of pseudopyloric glands in hyperplastic proliferation. An erosion was seen at the center of the lesion. the basal portion of the polyp is presumably pulled by the peristaltic movement of the stomach (Fig. 4). Polyps consisting of tnetaplastic epithelium occupied about 5% of all polyps. Sex ratio was 1:0.3. Multiple occurrence, which was 2 to 5 at most, was seen in 14% of 83 cases. The epithelial tubules in the polyps contain goblet cells and some times Paneth cells in a large number. The presence of brush border stained by alcian blue is also a feature of the intestinal epithelium. In most of them, marked atypism was seen in the superficial portion of the lesion, and was sometimes not readily differentiated from tubular adenocarcinoma (Fig. 5). The dense cells constituting the epithelial tubules have tall columnar cytoplasm with markedly increased eosinophilia, in which the amount of mucin content was reduced in inverse proportion to the increase of atypism (Fig. 6). The nuclei are short and spindle-shaped, pleomorphic and hyperchromatic with abundunt mitosis, having sometimes lost their basal orientation (Fig. 7). In our experience, carcinoma was found in 13% of the cases with this type of polyp, either in the same lesion or in

5 Gastric Polyp 27 Fig. 3. Type C of gastritis polyposa. (a) Gastrocamera (GTF) picture reveals multiple gastric polyps on the anterior wall of the lower portion of the corpus. A red erosion is seen on the top of each elevation. (b) Histologic picture shows elongated gastric foveolae consisting of regenerative indifferent epithelial cells due to erosion which was seen at the center of the picture. other parts of the same stomach. The commonest site of this type was the antrum where gastric mucosa showed severe atrophy with intestinal metaplasia. This type of polyps shows a low bank-like elevation with a broad base and a pale surface or a worm-like shape with unevenness of the surface, giving rise to a problem of differen tiation from early gastric cancer type IIa (Fig. 8). Among these polyps, 3 lesions were pedunculated. The epithelial cells in these polyps showed more marked atypism (Fig. 9). Very rarely, on the other hand, there were some polyps showing no or little atypism of cells and glands even in this type. Results of clinical observation In our mass-survey for gastric diseases, gastric polyp was found in 1,616 cases or 0.23% of 711,455 persons, with a yearly rate between 0.19% and 0.28%; on the other hand, gastric cancer was found in 0.2% in this series. The commonest site of these polyps was the antrum, occupying about 70% of all polyps. Females were more predominant than males with a ratio of 0.18:0.29 that is 1:1.6 (Table 2).

6 28 K. Ueno et al. Fig. 4. Type E of gasritis polyposa. (a) Resected stomach of a 19-year-old male showing multiple polyposis in the antrum. (b) and (c) Histologic pictures of sessile and pedunculated polyps of the same case, composed of both proliferated lymph follicles and hyperplastic epithelium. The survey of incidence of gastric polyp in different age groups revealed a tendency toward markedly increased gastric polyps with advancing age. When these polyps were classified into 4 gross pathological groups according to the shape of the lesion as shown in Fig. 10, sessile hemispherical polyp was most frequent in the young age group, whereas the increase of pedunculated polyp was most pronounced in older age groups. In 1,104 polyps, in which the follow-up observation has been done by means of

7 Gastric Polyp 29 Fig. 5. Histologic pictures of polyp composed of metaplastic epithelium showing atypism. In Japan, this type is called atypical epithelium (ATP). (a) A typical picture of this type found in Japan. (b) A part of the same lesion magnified. Fig. 6. Relationship between histologic atypism and the amount of cytoplasmic mucin contents of the atypical epithelium (ATP)., alcian-blue staining of cytoplasmic mucin (-);??, (+);??, (??). fiberscopy and biopsy during a time course of 6 months to 11 years at the longest, a growth in the size and in the figure was seen in 14 polyps, 12 out of which were hyperplastic polyps diagnosed by biopsy (Table 3). In the remaining two, one was a pedunculated one, probably hyperplastic polyp though no biopsy was carried out

8 30 K. Ueno et al. Fig. 7. Highly magnified histologic picture of ATP. Fig. 8. Endoscopic picture of ATP taken by GTF-S. during the period of observation; another was a polyp consisting of metaplastic epithelium. Since this case is not operated yet, no conclusion can be drawn unless pathological confirmation is made. The reason is that two other lesions, which had been followed up as belonging to this type and increased in size in the clinical course, were found to be carcinoma after operation. A change from gastritis polyposa into pedunculated polyp or an increase in size in gastritis polyposa was never experi enced. From these results, it can be said that most of the polyps with increase in size are hyperplastic polyps. In view of the frequency of 12 or 3.2% out of 380 hyperplastic polyps, however, the growth of hyperplastic polyp in the clinical course appears to be extremely rare. Most of hyperplastic polyps which

9 Gastric Polyp 3 Fig. 9. A case of ATP which has a pedicle (a) showing so marked atypism as neoplastic in nature (b). TABLE 2. Frequency of gastric polyp in different age groups

10 32 K. Ueno et al. Fig. 10. Relationship between age and type of gastric polyp., Type I;??, Type II;??, Type??;, Type IV. TABLE 3. Results of follow-up observation of gastric polyps Numbers in the parentheses show the grown-up polyp. increased in size were small sessile polyps with marked surface reddening, appear ing in the antrum (Table 4). The age of the patients averaged 53. The stage of the increase in size in 12 hyperplastic polyps was assigned to the beginning of the follow-up course in only one case, mostly to 3 to 4 years. The change in size was always transient, and a long-standing growth was never experienced. An increase in size seen in a few cases is, therefore, probably due to a secondary cause such as inflammatory change superimposed on these polyps. Fig. 11(a) shows a polyp in the antrum, and Fig. 11(b) shows the gastrocamera picture of the same case 2 years before. On comparison of these two pictures, a reddening with small elevation was seen in Fig. 11(b) corresponding to the place of the polyp in Fig. 11(a). Since we have experienced 7 such cases, some of gastric polyps might occur as a "polypbud" with a histologic feature under the red color tone similar to that of hyperplastic polyp from the beginning. After actual

11 Gastric Polyp 33 TABLE cases with hyperplastic polyp which increased in size during the follow-up period. Fig. 11. (a) A polyp in the antrum. (b) Gastrocamera picture of the same stomach taken 2 years before. observation of 12 such lesions newly detected, however, no tendency toward growth was noted in any case during the follow-up study of 1 to 3 years. Based on this observation, the period between the occurrence of polyp and the completion of its growth is rather short; and most of the polyps clinically detected have probably completed their growth even in such small polyps. The assumption was substantiated by the result of autoradiographic study using 3H-TdR; labeling index of epithelial cells composing hyperplastic polyp was nearly 0%, while labeling indices of adenocarcinoma, ATP and the mucosa of atrophic gastritis were 13.6%, 10.8% and 10.3%, respectively (Fig. 12).

12 34 K. Ueno et al. Fig. 12. Labeling index of epithelial cells in each gastric lesion (the result of autoradio graphic study using 3H-TdR). Relationship between gastric polyp and cancer Among 974 polyps in which biopsy had been done more than 2 times during the follow-up period, a change into cancer was experienced in none of the case. Among 297 polypoid lesions operated, which were histologically confirmed to be carcinoma in 61 and benign polyp in 236, 5 had cancer limited to only a small part of the protrusion. Histological feature of the other part of these lesions was similar to that of metaplastic polyp showing marked atypism in 2, and to that of hyperplastic polyp in 3. DISCUSSION As shown in Table 5 in which histological classifications of gastric polyp reported were listed, gastric polyp is grossly classified into 3 groups : a group of pseudopolyp or gastritis polyposa, a group of hyperplastic polyp, and a group called adenoma or metaplastic polyp (ATP). The former two groups, which were composed of gastric epithelium with or without some degree of hyperplasia, have been mostly considered to be the result of regeneration and are termed regenerative polyp by Ming and Goldman (1965). Hyperplastic polyp is, however, distinguished by marked mucosal proliferation with differentiated epithelial glands intermingled with collagenous and smooth muscle fibers originating from muscularis mucosae, while gastritis polyposa consists of 5 types of polyps which show each characteristic histo-clinical features. Type A is a rare polyp of unknown course, of which no report has ever been published. Types B and C, namely gastritis verrucosa, are the most

13 Gastric Polyp 35 TABLE 5. Histological classifications of gastric polyp. Types in these classifications which show common characteristics are arranged in the same column. Types with mixed characteristics of two types are put between the columns of the two. * A type in a different sense from other types common types and are just considered to be the result of regeneration. These must be distinguished from verrucous protrusion of gastritis erosiva due to in flammatory edema. The histologic feature, particularly of type C, is in continuity with that of hyperplastic polyp. In reality, coexistence of pedunculated hyperplastic polyp and gastritis verrucosa in the same stomach was found in 9.5% in our experience and in 11% of cases of this type by Takasu et al. (1974). Based on these facts, a possibility of growth from gastritis verrucosa to pedunculated hyperplastic polyp has been suggested (Sano et al. 1970). In our results, however, no change from this type into hyperplastic polyp or an increase in size of polyps in gastritis polyposa was experienced. Type D is just a polyp called pseudopolyp. Type E is a special type composed of both hyperplastic epithelia and proliferated lymph follicles. For the study of gastric polyp, therefore, it will be necessary to investigate each type of gastritis polyposa as well as other polyps. Since Carey and Hay (1948) stated, based on the result of their follow-up study, that there was no direct relationship between gastric carcinoma and gastric polyp, gastric polyp is considered mostly to be benign. Yarnis et al. (1952) found polyp cancer in 2 or 6.7% out of 30 cases, Sano (1968) in 5 or 7% of 69 pedunculated polyps, Sugano et al. (1968) in 5 or 9% of 53, and Nakamura (1968) in 3 or 4.8% of 62 polyps. In our result, malignant degeneration was suspected in 5 out of 61 polypoid cancers and 236 gastric polyps operated; that was 2% of 236 gastric polyps. The evidence of malignant degeneration in gastric polyp is that carcinoma is limited to a small part of the lesion showing mostly histologic features of benign gastric polyp. Even in this criterion, however, the original nature of these lesions cannot be known definitely; whether they were gastric polyp undergoing malignant degeneration or carcinoma from the beginning is uncertain. In reality, no case of polyp which changed into cancer was experienced in 1,016 polyps followed-up by biopsy.

14 36 K. Ueno et al. Since the existence of gastric polyp composed of metaplastic epithelium was first pointed out with the special reference to gastric carcinoma by Morson (1955), this type of polyp has been separated from other polyps by Nakamura (1964) and by Ming and Goldman (1965). In Japan, this type is called ATP (atypical epithelium), and attention is focused on its relationship with cancer (Nakamura et al. 1966). Actually, coexistence with gastric cancer was found in 40% by Ming, in 75% by Berg (1958), in 15.6% by Nakamura and in 13.2% in our experience. In foreign countries, these are considered mostly to be neoplastic in nature for their marked atypism and for their special reference to carcinoma, and are called adenomatous polyp (Ming and Goldman 1965; Evans 1966) or adenoma (Morson et al. 1972a; Debray and Martin 1974), like adenoma of colonic polyp (Morson et al. 1972b). However, it is really difficult to know whether each of this type is neoplastic in nature or hyperplastic reaction in origin. The reason for this is twofold. First, no tendency of the growth was seen in this group of polyps during the clinical course of our follow-up observation. The growth of a neoplastic lesion may persist in the same excessive manner in the clinical course though it is not necessarily seen (Meissner and Warren 1971). Second, all grades of atypism were seen in every polyp of this type, showing histological continuity with one another; a polyp showing little atypism with many goblet cells, Paneth cells and a large mucin content in the cytoplasm may be safely said to be no neoplastic in nature, while a polyp showing marked atypism with the reduced mucin content in the cytoplasm is not easily differentiated from carcinomạ The latter one might be called adenoma, if this could be exactly distinguished from carcinoma in situ or well differentiated tubular adenocarcinoma. In general, most of polyps composed of the metaplastic epithelium detected in our country revealed low flower bed-like elevation with a broad base, mostly within 2 cm in size, in which atypical epithelium was limited to the surface and the foveolar portion of the glands (Sugano et al. 1971). Sugano et al. (1971) suggested two possibilities for the outcome of these atypi cal epithelium: differentiation to intestinal metaplasia and transformation into cancer. These low flower bed-like lesions may correspond to those described by Ming and Goldman (1965) as "earlier lesion". On the other hand, 40% of adenomatous polyps reported by Ming and Goldman (1965) were pedunculated and 80% of them were those over 2 cm in size. The reason for this discrepancy between our country and foreign countries may be threefold. Firstly, the incidence of the lesion may be different in different races. Secondly, many small lesions as below 1 cm in size have been found clinically by routine endoscopic examination as well as biopsy in Japan. Thirdly, the histologic criterion for malignancy varies from one author to another and a consensus is not arrived at in this respecṭ Recently, Nakamura (1972) made a change on his classification reported in 1964, adding a new type which is pedunculated polyp showing more marked atypism and is considered to be neoplastic in nature. This type is rare in our country and 3 or 3.6% out of 83 cases in our series. Based on these facts, it can be said that there are probably two kinds of polyps in ATP: a metaplastic polyp in hyperplastic nature and the

15 Gastric Polyp 37 other neoplastic in nature. It has been known that the incidence of gastric polyp, especially of pedunculated polyp, reveals a tendency of increase with increasing age. However, no report is available on whether and how a small polyp really grows into a large one with advancing age. From the results of our follow-up study, it can be said that most of gastric polyps clinically detected have already completed their growth even in small polyps. The period between the occurrence of gastric polyp and the completion of its growth is rather short and every polyp reaches the different predestined final size in a short time after the appearance, probably within 3 months, and scarcely changes thereafter. Even if an increase in the size may be seen in a few cases (3% or so), it is probably due to a secondary cause superimposed upon the polyp. References 1) Abel, W. (1954) Die R0ntgendiagnose der Gastritis erosiva. Fortschr. Rontgenstr., 80, ) Berg, J.W. (1958) Histological aspect of the relation between gastric adenomatous polyps and gastric cancer. Cancer, 11, ) Carey, J.B. & Hay, L.J. (1948) Gastric polyps. Gastroenterology, 14, ) Castleman, B. & Krickstein, H. (1966) Current approach to the polyp-cancer controversy. Gastroenterology, 51, ) Debray, C. & Martin, E. (1974) Benign gastric tumors. In: Gastroenterology. 3rd ed., edited by H.L. Bockus, W.B. Saunders, Philadelphia-London-Toronto, pp ) Evans, W. (1966) Histological Appearances of Tumors. E.S. Livingstone, Edinburgh London, pp ) Konjetzny, G.E. (1928) Die ntziundungen des Magens. Henke-Lubarsch's Handbuch d. spez. pathol. Anatomic and Histologie, Bd., IV, 2, Springer Verlag, Berlin, pp ) Maruyama, M., Ichioka, S., Takemoto, T., Kondo, T., Yamagata, S., Ueno, K., Tsuneoka, K. & Sesoko, M. (1972) A new type of fibergastroscope (FGS-BLG) for measurement of gastric lesions. Tohoku J. exp. Med., 107, ) Meissner, W.A. & Warren, S. (1971) Neoplasms. Pathology. 6th ed., edited by W.A.D. Anderson & C.V. Mosby, St. Louis, p ) Ming, S. & Goldman, H. (1965) A histogenetic classification and its relation to carci noma. Cancer, 18, ) Monaco, A.P., Roth, S.I., Castleman, B. & Welch, C.E. (1962) Adenomatous polyps of the stomach; a clinical and pathological study of 153 cases. Cancer, 15, ) Morson, B.C. (1955) Gastric polyps composed of intestinal epithelium. Brit. J. Cancer, 8, ) Morson, B.C. & Dawson, I.M.P. (1972a) Gastrointestinal Pathology, Blackwell Scientific Publications, Oxford-London-Edinburgh-Melbourne, pp ) Morson, B.C. & Dawson, I.M.P. (1972b) Gastrointestinal Pathology, Blackwell Scientific Publications, Oxford-London-Edinburgh-Melbourne, pp ) Nakamura, T. (1964) Pathology of gastric polyp. Igaku no Ayumi (Jap.), 49, ) Nakamura, T. (1968) Malignant degeneration of gastric polyp; reference to his topathological classification. Stomach and Intestine (Jap.), 3, ) Nakamura, T. (1972) Gastric polyp. Early gastric cancer. Internal Medicine, Series No. 8 (Jap.), edited by K. Tsuneoka, Nankodo, Tokyo, pp ) Nakamura, K., Sugano, H., Takagi, K. & Fuchigami, A. (1966) Histopathological study of early carcinoma of the stomach; criteria for diagnosis of atypical epithelium.

16 38 K. Ueno et al. Gann, 57, ) Sano, R. (1968) Relationship between gastric adenomatous polyps and its malignant potential. Stomach and Intestine (Jap.), 3, ) Sano, R., Hirota, E. & Shimoda, T. (1970) Pathological study of early gastric cancer; relationship between cancer and gastritis verrucosa. Trans. Soc. path. Jap. (Jap.), 59, ) Schindler, R. (1950) Gastroscopy; the Endoscopic Study of Gastric Pathology. 2nd ed., The University of Chicago Press, Chicago, pp ) Stout, A.P. (1953) Tumors of the stomach. Atlas of Tumor Pathology, Sec. VI, Armed Forces Institute of Pathology, Washington, D.C., pp ) Sugano, H., Nakamura, K. & Takagi, K. (1968) Pathological study of gastric polyp, polyp-cancer and polypoid cancer. S7omach and Intestine (Jap.), 3, ) Sugano, H., Nakamura, K. & Takagi, K. (1971) An atypical epithelium of the stomach. Gann Monograph on. Cancer Research, vol. 11, University of Tokyo Press, Tokyo, pp ) Tachadjian, V. (1963) Benign tumors of the stomach. Gastroenterology. 2nd ed., Vol. 1, edited by H. Bockus, W.B. Saunders, Philadelphia-London, pp ) Takasu, S., Inui, Y. & Garcia, J.C. (1974) A report of 100 cases of gastritis verrucosa and their long-term follow-up. Gastroent. Endoscopy (Jap.), 16, ) Yarnis, H., Marshak, R.H. & Friedman, A.I. (1952) Gastric polyps. J. Amer. med. Ass., 148,